Confirmatory Factor Analysis of the Parental Attitudes, Beliefs, and Understanding of Anxiety (PABUA)

The Parental Attitudes, Beliefs, and Understanding of Anxiety (PABUA; Wolk et al., 2016) was developed to assess caregivers’ beliefs about, perceived ability to tolerate or manage, and understanding of strategies for responding to their child’s anxiety. Following a preliminary validation of the PABUA that involved exploratory factor analysis (Wolk et al., 2016), this study aimed to confirm the factor structure in an independent sample and to further evaluate concurrent and discriminant validity of the subscales: (1) Overprotection, or beliefs that youth should be protected from feeling anxious; (2) Distress, or appraisals of guilt, anxiety/discomfort, helplessness, and avoidance in response to youth anxiety; and (3) Approach, or the belief in the value of promoting autonomy and modeling for/encouraging youth to engage with anxiety-provoking situations. Participants (N = 207; 70% female; 87.4% White; 91.8% non-Hispanic) were caregivers of a child 6 to 17 years old with clinical levels of anxiety (based on SCARED-P score 25) and were recruited via Amazon Mechanical Turk. They were administered the PABUA, a measure of experiential avoidance in the parenting context, a measure of the belief that youth anxiety is harmful, and self-report measures of anxiety and depression. Results of confirmatory factor analysis with diagonally least weighted squares estimation supported the initial three-factor model, with some items removed (model fit indices: CFI: .98; TLI: .98; RMSEA: .06; SRMR: .08). There were 8 items on Overprotection ( = .88), 5 items on Distress ( = .66), and 3 items on Approach – which did not meet the standard for acceptable internal consistency. There was also support for a two-factor model that excludes the Approach scale (model fit indices: CFI: .99; TLI: .99; RMSEA: .05; SRMR: .07), with improved internal consistency for the Distress subscale (Overprotection:  = .85; Distress:  = .68). Correlational analysis to examine concurrent and discriminant validity generally replicated findings from the initial validation study. For example, using scales based on the two-factor model, PABUA Overprotection was more strongly associated with a measure of unwillingness to experience a child’s negative emotion (r = .50, p < .01) than was PABUA Distress (r = .27, p <.05); whereas PABUA Distress was more strongly associated with a measure of caregiver inability to manage their reaction to a child’s negative emotion (r = .61, p < .01) than was PAUBA Overprotection (r = .32, p < .01). The PABUA can be used to study the impact of caregiver beliefs on a child’s anxiety and may allow for systematic assessment of family-based treatment targets. Efforts to refine the PABUA should focus on expanding the Approach subscale and reconsidering item anchors to increase variability in responses.